This project aims to develop statistical and computational methods for Identifying and haracterizing SNPs (genes) and environmental factors in the involvement of disease occurrence and progression using state-of-art data that are recently generated by high throughput technologies. The first aim proposes to develop and study likelihood based regression strategies to construct flexible combinations of environmental factors that modify genetic effects in association studies that include binary, continuous and time-to-event data. Boosting and regularized regression strategies are used in structured and unstructured interaction models. The second aim proposes the development of improved test statistics for assessing the association of rare variants with disease risk in sequencing studies by adaptively selecting variants variants to use, and by incorporating genome-wide association studies on additional subjects who are not sequenced to increase power. A third aim is concerned with methods development for estimating age-specific absolute risk of genetic variants and environmental factors from case-control studies, which includes a flexible and efficient estimation for time-varying attributable risk function critical in obtaining unbiased estimators of absolute risk and semi- and non-parametric estimation of composite incidence rate from the family history data on cases and controls. A final aim will develop methods to test and estimate the causal effect of an exposure on a clinical outcome, using genetic variants as instruments. The methods exploit the Mendelian randomization of genetic variants and the dose correspondence of genetic effects on the exposure and the the outcome. The project draws on the strength of the studies in which the five investigators are directly involved and addresses the needs and barriers that these studies face. The methods and tools developed in this project will have a broad application to large-scale genetic epidemiologic studies.